3 This directory contains the test suite for notmuch.
5 When fixing bugs or enhancing notmuch, you are strongly encouraged to
6 add tests in this directory to cover what you are trying to fix or
11 The test system itself requires:
13 - bash(1) version 4.0 or newer
15 Without bash 4.0+ the tests just refuse to run.
17 Some tests require external dependencies to run. Without them, they
18 will be skipped, or (rarely) marked failed. Please install these, so
19 that you know if you break anything.
29 If your system lacks these tools or have older, non-upgradable versions
30 of these, please (possibly compile and) install these to some other
31 path, for example /usr/local/bin or /opt/gnu/bin. Then prepend the
32 chosen directory to your PATH before running the tests.
34 e.g. env PATH=/opt/gnu/bin:$PATH make test
36 For FreeBSD you need to install latest gdb from ports or packages and
37 provide path to it in TEST_GDB environment variable before executing
38 the tests, native FreeBSD gdb does not not work. If you install
39 coreutils, which provides GNU versions of basic utils like 'date' and
40 'base64' on FreeBSD, the test suite will use these instead of the
41 native ones. This provides robustness against portability issues with
42 these system tools. Most often the tests are written, reviewed and
43 tested on Linux system so such portability issues arise from time to
49 The easiest way to run tests is to say "make test", (or simply run the
50 notmuch-test script). Either command will run all available tests.
52 Alternately, you can run a specific subset of tests by simply invoking
53 one of the executable scripts in this directory, (such as ./T*-search.sh,
54 ./T*-reply.sh, etc). Note that you will probably want "make test-binaries"
55 before running individual tests.
57 The following command-line options are available when running tests:
60 This may help the person who is developing a new test.
61 It causes the command defined with test_debug to run.
64 This causes the test to immediately exit upon the first
68 Execute notmuch with valgrind and exit with status
69 126 on errors (just like regular tests, this will only stop
70 the test script when running under -i). Valgrind errors
71 go to stderr, so you might want to pass the -v option, too.
73 Since it makes no sense to run the tests with --valgrind and
74 not see any output, this option implies --verbose. For
75 convenience, it also implies --tee.
78 In addition to printing the test output to the terminal,
79 write it to files named 't/test-results/$TEST_NAME.out'.
80 As the names depend on the tests' file names, it is safe to
81 run the tests with this option in parallel.
83 Certain tests require precomputed databases to complete. You can fetch these
86 make download-test-databases
88 If you do not download the test databases, the relevant tests will be
91 When invoking the test suite via "make test" any of the above options
92 can be specified as follows:
94 make test OPTIONS="--verbose"
96 You can choose an emacs binary (and corresponding emacsclient) to run
97 the tests in one of the following ways.
99 TEST_EMACS=my-emacs TEST_EMACSCLIENT=my-emacsclient make test
100 TEST_EMACS=my-emacs TEST_EMACSCLIENT=my-emacsclient ./T*-emacs.sh
101 make test TEST_EMACS=my-emacs TEST_EMACSCLIENT=my-emacsclient
103 Some tests may require a c compiler. You can choose the name and flags similarly
106 make test TEST_CC=gcc TEST_CFLAGS="-g -O2"
111 Normally, when new script starts and when test PASSes you get a message
112 printed on screen. This printing can be disabled by setting the
113 NOTMUCH_TEST_QUIET variable to a non-null value. Message on test
114 failures and skips are still printed.
118 If, for any reason, you need to skip one or more tests, you can do so
119 by setting the NOTMUCH_SKIP_TESTS variable to the name of one or more
124 $ NOTMUCH_SKIP_TESTS="search reply" make test
126 Even more fine-grained skipping is possible by appending a test number
127 (or glob pattern) after the section name. For example, the first
128 search test and the second reply test could be skipped with:
130 $ NOTMUCH_SKIP_TESTS="search.1 reply.2" make test
132 Note that some tests in the existing test suite rely on previous test
133 items, so you cannot arbitrarily skip any test and expect the
134 remaining tests to be unaffected.
136 Currently we do not consider skipped tests as build failures. For
137 maximum robustness, when setting up automated build processes, you
138 should explicitly skip tests, rather than relying on notmuch's
139 detection of missing prerequisites. In the future we may treat tests
140 unable to run because of missing prerequisites, but not explicitly
141 skipped by the user, as failures.
145 The test script is written as a shell script. It is to be named as
146 Tddd-testname.sh where 'ddd' is three digits and 'testname' the "bare"
147 name of your test. Tests will be run in order the 'ddd' part determines.
149 The test script should start with the standard "#!/usr/bin/env bash"
150 and an assignment to variable 'test_description', like this:
154 test_description='xxx test (option --frotz)
156 This test exercises the "notmuch xxx" command when
157 given the option --frotz.'
161 After assigning test_description, the test script should source
162 test-lib.sh like this:
164 . ./test-lib.sh || exit 1
166 This test harness library does the following things:
168 - If the script is invoked with command line argument --help
169 (or -h), it shows the test_description and exits.
171 - Creates a temporary directory with default notmuch-config and a
172 mail store with a corpus of mail, (initially, 50 early messages
173 sent to the notmuch list). This directory is
174 test/tmp.<test-basename>. The path to notmuch-config is exported in
175 NOTMUCH_CONFIG environment variable and mail store path is stored
176 in MAIL_DIR variable.
178 - Defines standard test helper functions for your scripts to
179 use. These functions are designed to make all scripts behave
180 consistently when command line arguments --verbose (or -v),
181 --debug (or -d), and --immediate (or -i) is given.
185 Your script will be a sequence of tests, using helper functions
186 from the test harness library. At the end of the script, call
191 There are a handful helper functions defined in the test harness
192 library for your script to use.
194 test_begin_subtest <message>
196 Set the test description message for a subsequent test_expect_*
197 invocation (see below).
199 test_expect_success <script>
201 This takes a string as parameter, and evaluates the
202 <script>. If it yields success, test is considered
205 test_expect_code <code> <script>
207 This takes two strings as parameter, and evaluates the <script>.
208 If it yields <code> exit status, test is considered successful.
210 test_subtest_known_broken
212 Mark the current test as broken. Such tests are expected to fail.
213 Unlike the normal tests, which say "PASS" on success and "FAIL" on
214 failure, these will say "FIXED" on success and "BROKEN" on failure.
215 Failures from these tests won't cause -i (immediate) to stop. A
216 test must call this before any test_expect_* function.
218 test_expect_equal <output> <expected>
220 This is an often-used convenience function built on top of
221 test_expect_success. It uses the message from the last
222 test_begin_subtest call, so call before calling
223 test_expect_equal. This function generates a successful test if
224 both the <output> and <expected> strings are identical. If not, it
225 will generate a failure and print the difference of the two
228 test_expect_equal_file <file1> <file2>
230 Identical to test_expect_equal, except that <file1> and <file2>
231 are files instead of strings. This is a much more robust method to
232 compare formatted textual information, since it also notices
233 whitespace and closing newline differences.
235 test_expect_equal_json <output> <expected>
237 Identical to test_expect_equal, except that the two strings are
238 treated as JSON and canonicalized before equality testing. This is
239 useful to abstract away from whitespace differences in the expected
240 output and that generated by running a notmuch command.
244 This takes a single argument, <script>, and evaluates it only
245 when the test script is started with --debug command line
246 argument. This is primarily meant for use during the
247 development of a new test script.
249 test_emacs <emacs-lisp-expressions>
251 This function executes the provided emacs lisp script within
252 emacs. The script can be a sequence of emacs lisp expressions,
253 (that is, they will be evaluated within a progn form). Emacs
254 stdout and stderr is not available, the common way to get output
255 is to save it to a file. There are some auxiliary functions
256 useful in emacs tests provided in test-lib.el. Do not use `setq'
257 for setting variables in Emacs tests because it affects other
258 tests that may run in the same Emacs instance. Use `let' instead
259 so the scope of the changed variables is limited to a single test.
261 test_emacs_expect_t <emacs-lisp-expressions>
263 This function executes the provided emacs lisp script within
264 emacs in a manner similar to 'test_emacs'. The expressions should
265 return the value `t' to indicate that the test has passed. If the
266 test does not return `t' then it is considered failed and all data
267 returned by the test is reported to the tester.
271 Your test script must have test_done at the end. Its purpose
272 is to summarize successes and failures in the test script and
273 exit with an appropriate error code.
275 There are also a number of notmuch-specific auxiliary functions and
276 variables which are useful in writing tests:
280 Generates a message with an optional template. Most tests will
281 actually prefer to call add_message. See below.
285 Generate a message and add it to the database (by calling "notmuch
286 new"). It is sufficient to simply call add_message with no
287 arguments if you don't care about the content of the message. If
288 more control is needed, arguments can be provide to specify many
289 different header values for the new message. See the documentation
290 within test-lib.sh or refer to many example calls within existing
295 This function should be called at the beginning of a test file
296 when a test needs to operate on a non-empty body of messages. It
297 will initialize the mail database to a known state of 50 sample
298 messages, (culled from the early history of the notmuch mailing
301 notmuch_counter_reset
302 $notmuch_counter_command
303 notmuch_counter_value
305 These allow to count how many times notmuch binary is called.
306 notmuch_counter_reset() function generates a script that counts
307 how many times it is called and resets the counter to zero. The
308 function sets $notmuch_counter_command variable to the path to the
309 generated script that should be called instead of notmuch to do
310 the counting. The notmuch_counter_value() function prints the
311 current counter value.
313 There are also functions which remove various environment-dependent
314 values from notmuch output; these are useful to ensure that test
315 results remain consistent across different machines.
317 notmuch_search_sanitize
318 notmuch_show_sanitize
319 notmuch_show_sanitize_all
320 notmuch_json_show_sanitize
322 All these functions should receive the text to be sanitized as the
323 input of a pipe, e.g.
324 output=`notmuch search "..." | notmuch_search_sanitize`